Higgsfield Raises $400M Series B at $5.4B Valuation

Higgsfield has raised a $400M Series B at a $5.4B post-money valuation, according to Axios. DST Global led the financing, joined by a broad group of new and returning investors spanning venture capital, growth equity, corporate venture, media, and telecommunications.

The round matters because Higgsfield is not being priced as a clever consumer video app. Investors are underwriting a larger thesis: the workflow layer around generative video can become durable infrastructure for commercial creativity. Higgsfield builds its own creative tools, integrates outside models, and packages that moving technical stack into workflows that creators and marketing teams can actually use.

The valuation also compresses a remarkable amount of company-building into a short period. Higgsfield was founded in 2023, launched its browser product in 2025, and announced an $80M Series A extension in January 2026 that brought total Series A financing above $130M at a valuation above $1.3B. Seven months later, the post-money valuation is $5.4B.

What Happened

DST Global led the $400M financing. The reported syndicate includes Tribe Capital, Goldman Sachs Alternatives, Smash Capital, Fifth Wall, Valor Capital, Intel Capital, Liberty Global Tech Ventures, Mirae Asset Capital, and NTT Docomo Ventures. Returning investors Accel, Menlo Ventures, AI Capital Partners, GFT Ventures, Capra Ventures, BAM Corner Point, and BroadLight Capital also participated.

The Series B follows an unusually fast financing history. Menlo Ventures led Higgsfield's $8M seed round in April 2024. A $50M Series A followed in 2025, and the January 2026 extension added $80M. With the current financing, Higgsfield has disclosed more than $538M across those rounds.

The exact use of proceeds for the Series B was not disclosed in the current reporting reviewed by DevCuration. That matters because the January financing came with specific plans around enterprise expansion, international growth, research, APIs, and marketing automation. Those prior plans provide context, but they should not be recycled as the official mandate for a new $400M round.

Why Higgsfield Looks Like Infrastructure

Higgsfield's product strategy is built around abstraction. The company develops proprietary systems such as Soul 2.0, Soul Cinema, Higgsfield DOP, Keyframes, Cinema Studio, Soul ID, and Soul HEX, while also integrating external models into one production environment. The user does not have to rebuild an entire creative pipeline every time a model improves, disappears, changes price, or becomes better suited to a particular task.

That workflow layer is where application companies try to create defensibility. Access to a leading model can be rented. A repeatable system for ideation, storyboarding, generation, editing, consistency, collaboration, and delivery is harder to replace once a team has built real operating habits around it. Reuters described Higgsfield's approach as a reasoning engine that chains multiple AI systems to support consistent characters and branding.

For buyers, the value proposition is less philosophical. Marketing teams need more video, more variants, faster turnaround, and enough control to avoid turning every campaign into an expensive experiment. If Higgsfield can make generative production predictable for nontechnical teams, the company earns its place between model providers and the budgets already moving through agencies, brand studios, and enterprise creative departments.

The Operating Numbers Behind the Round

Higgsfield's current company profile says the platform serves 25M users who have produced 850M generations, including 300M videos. In June 2026, the company told Business Insider that it had reached a $500M annualized revenue run rate, was cash-flow positive at the end of May, worked with 390 Fortune 500 companies, and generated 70% of platform activity from commercial advertising.

Those figures are company-reported. They are not the same thing as audited revenue, and run rate is not recognized revenue. Reuters made that distinction explicit when Higgsfield reported a $200M annualized run rate in January. The qualification does not make the numbers irrelevant. It makes them the beginning of the diligence conversation rather than the end.

Investors appear to be betting that the velocity reflects real workflow adoption rather than a temporary burst of consumer curiosity. The breadth of the syndicate is notable here. Financial investors are joined by strategic firms connected to chips, telecom, media, and distribution, all markets with a reason to care about how synthetic media is created, financed, governed, and delivered.

The Leadership and Market Bet

Accel identifies Alex Mashrabov, Yerzat Dulat, and Mahi de Silva as Higgsfield's founders. Mashrabov serves as CEO after previously leading generative AI work at Snap and co-founding AI Factory. Dulat is the company's CTO, while de Silva is Chief Strategy Officer. That combination gives Higgsfield technical, product, media, and go-to-market experience at a moment when the category needs all 4 at once.

The market bet is bigger than whether AI can produce a convincing clip. That argument is already old. The useful question is whether companies can build trusted, controllable, and economical production systems around models that remain expensive, fast-changing, and dependent on outside infrastructure.

Higgsfield also has to manage the tension inside its own strategy. Supporting multiple models gives customers choice and protects against betting everything on one provider. It can also create supplier dependency, shifting economics, governance complexity, and the constant burden of making inconsistent systems feel like one coherent product.

What This Signals for AI Video

The $400M Series B says investors see meaningful value above the foundation-model layer. That is the same basic wager behind successful application companies in coding, customer support, legal work, and design: customers pay for outcomes, workflow, reliability, and control, not for a raw technical benchmark.

For creative-software incumbents, Higgsfield's rise is a warning about product tempo. A company that can integrate new models quickly, package them into usable workflows, and learn from a large volume of production activity may move faster than a suite built around older editing assumptions. For model providers, the round is evidence that distribution and workflow partners can be as important as direct consumer products.

The next proof points are less cinematic and more operational. Higgsfield has to convert self-reported run-rate growth into durable revenue, keep enterprise users after initial experimentation, manage compute economics, strengthen governance, and preserve creative control as output volume grows. The financing buys scale, but it also raises the standard by which that scale will be judged.

Higgsfield's founders deserve credit for turning a young product into one of the largest generative-media financings on record. The harder accomplishment begins now: proving that AI video can behave like dependable production software without stripping away the taste, judgment, and accountability that make commercial creativity valuable.